I wanted to explore a few additional details, create a historical record for posterity and completely debunk the 1,300 tonnes of leased gold idea before it gains too much traction in the metals space. Primarily though, the main purpose of this article is to highlight extra detail - insofar as it helps reduce the social distance between us and that big stash of gold.
|Worth the visit - an interactive gold vault complete with one-tonne shelves.|
"The Royal Visit took place in our display vault. This is the vault used in the Bank’s official photographs and is the image the general public are familiar with from archive footage. It is distinctive..." -- Louise Lee, Manager of Gold and Foreign Exchange Settlement Operations, The Bank of England (link)On each shelf in the display vault we see in the picture is approximately one tonne, typically 80 bars in 4 layers of 20. The gold is stacked 4 shelves high, i.e. 4 tonnes = 320 bars in each vertical space. My count for the display vault from the virtual tour photo is 172 stacks, i.e. approximately 55,040 bars [~688 tonnes]. The amount here is difficult to conceptualize by volume, but by way of comparison consider that GLD at its peak once held 109,318 bars (but now less than 75,000).
In the professor's documentary, the stacks are 6 shelves high and based on the contents, layout, ceiling and floor markings, the professor's room is different to the main display vault. For the section visible to the camera (and inferred from rows continuing), I count 68 stacks, approximate total of 32,720 bars [~409 tonnes]. I've done my best to stitch together a floor layout of the sections which are known. Due to a sloping wall section which seems to be common to both, I've surmised that the professor's room adjoins the display vault (this may or may not be correct, it just seems to fit). Click on the image to see all my notes in detail. Feel free to comment with observations of your own.
So we know The Queen's visit took place in the display vault, and we can confirm this from matching the arrangement of columns supporting the roof and other clues*, but the most notable observation is that during the royal visit there are fewer gold stacks present compared with the virtual tour picture. There are of course a multitude of reasons to explain the discrepancy, some which are blindingly obvious - it would have been useful to fill the room completely with gold specifically for the photo-shoot to maximize the visual impact, or perhaps gold was moved out during the royal visit simply to make room for all the people in the room. Either way, we have no way of telling what the amount is 'normal', except to point out that it it generally 'full' during documentaries like David Dimbleby's brief visit and other photo shoots. The Alchemist extract claims the room holds 63,000 gold bars, so that is currently the most authoritative figure for the room contents.
My theory is that the 360-degree picture was taken prior to December 2012, because in the virtual tour picture, you can spin the camera around to view the ceiling you'll notice a few wires sticking out - I'm guessing this is the initial installation of additional security cameras which are visible on the roof during the Queens visit, but it may also be possible the cameras we see during the Queen's visit might have been taken out (or replaced) later with only wires remaining.
|Snapshot of my roof inspection comparison. Note it's still possible the supporting|
columns are hiding some of the detail, but generally the Virtual Tour photo
has a roof is devoid of cameras, where as many can be seen in the royal visit.
Why is the timing relevant? Over the last few days there's been some discussion about exactly how much gold is actually being held there, with some speculating that the differences between the estimates may be indicative of leasing arrangements (link). I find this incredibly unlikely because of the inherent 'limit of reading error' on all the figures available, and in particular the ones being used by Macleod/Keiser. They need to know that although the virtual tour was released in June/July of 2013, the production and construction of the website would have taken longer than that. Software projects invariably take many months to build and sometimes even longer to review and sign off before going 'to production' - the audience for the content is also vastly different to the official report! As such, the figure of 'over 400,000' gold bars should be taken for what it is - a round figure supplied for mass public consumption carefully weighed by legalese with enough ambiguity to be approximately correct at all times, rather than being an official statement about inventories. As such, Max Keiser, Alisdair Macleod as well as Jesse and anyone else caught perpetuating the story ought to have known better but perhaps they can be forgiven for not knowing the inane complexities of modern software development.
Where does that leave us? Our full tally for rooms on camera (with the assumption the inventory shown in each is unique) is at least 87,760 bars (~1097 tonnes). If we use 400,000 as the benchmark, this implies we've seen less than 21% of the inventory the bank holds, so therefore there must be more sections of the vault. Is there any detail on this? Yes. The virtual tour claims the bars are spread over 'two floors', but conducting a basic internet search reveals other snippets about the vault levels (my emphasis):
".. this vast building now rises seven storeys above ground level and has a further three basement levels, which contain the cavernous bullion vaults" linkI've also seen other references to the Display Vault being a former 'canteen', I think its fair to suggest if we were shown the rest of the vault space we would see the 4½m x 6m rooms mentioned above, and the big rooms we see on camera were re-purposed for gold storage at some point in history. The claim about abundant vault floorspace is also interesting, but not surprising if you consider that some additional £489 billion worth of 'assets' are also being stored¹.
"The vaults are huge and include three disused wells. In fact, the floorspace is actually larger than that of the City's tallest building, Tower 42. Keys three-feet long are needed to open the gigantic vault doors." link
"The basement levels contain the secure vaults, divided into strong rooms typically 6.10m by 4.57m, and separated by 610mm thick reinforced concrete walls and floors with additional armour plating. All the vaults are also watertight." link
For anyone doing their own research out there, here's my summary collation of the different figures quoted in relation to the BoE gold holdings². The numbers varied so much that I decided to add my own commentary on each. The figures used for my conversions are in this spreadsheet.
the Alchemist special feature from Louise Lee claims the Display Vault holds approximately 63,000 bars (784 tonnes), just 12% over my estimate of a full room. As such, this must represent the peak capacity, and not the actual contents when monarchs visit.
link). This series was broadcast in 2010, using average prices for the period gives a rough estimate of 3,500 tonnes of gold.
link), explains the total hoard is worth £197 billion. For the gold price at estimated Nov/Dec 2012, that would make the total weighing in at 5,836 tonnes. The professor does explain the figure is from the banks annual report but doesn't say which one so the result is blurry.
The Bank of England report published June 2013, is the one used by Alisadair and Max to come up with their leasing conclusion, but the figure is only in pounds, not ounces. They translate '£210 billion in gold as at 28th February 2013' as being 505,000 bars -> my own calculation is approximately 6,266 tonnes (503,637 bars).
Newspapers showed us some of the very first pictures of the BoE Display Vault, The Daily Mail in October 2008 quotes a figure of 4,600 tonnes (link), however it's not clear where they got that value from.
Because of the ambiguity in the figures it is possible to select any number or situation and use it to support a narrative. For example, I could use David Dimbleby's figures to argue that the bank has added thousands of tonnes to its holdings over the last few years - or even take the view that the display vault being full of gold in the Virtual Tour application in July 2013 indicates more gold had been added to the room since December 2012. Both views would be difficult to prove or disprove and neither have a specific bearing on gold prices past and future.
Finally it is worth noting the Macleod/Keiser theory absolutely rests on the idea that the app took LESS than 5 months to develop and launch to production, i.e. if the app was designed and copy provided BEFORE their end-of-February amount, then the same argument method means the bank ADDED more gold — the opposite thesis of what they are setting out to prove. So when I read "The app has been under development for about a year, a Bank spokeswoman said" (link), this is a SERIOUS BLOW to their leasing theory (however of course, the text may have been changed through a change request)**.
The moral of the story is be careful with your initial data points because it cuts both ways if you're reaching conclusions using disparate information.
* Other visual clues include the yellow painted section at one end as well as the shape of the covers on the cabling near doors (yes, I did quite a lot of study on the videos and pictures).
1. According to the same 2013 report. £489 billion is a lot of assets, and I've wondered what that is exactly. In internal discussion, the closest candidate is 'vault cash', but that still seems excessive even for a money center like London - perhaps someone with expertise in M0, MB, M1, M2, etc. can shed some light on that.
2. As far as I know, some of this Bank of England gold belongs to GLD shareholders, although to date we have never identified a GLD bar in any pictures (using the database). This actually makes perfect sense (no identified matches) in the light of the very large amounts (sample size) they have in their vaults. [Addendum: 01-August-2013, the use of BoE as a sub-custodian is stated here, the text suggests that any GLD gold allocated from BoE vault will be transferred to HSBC's main vault, i.e. not residing in the BoE vault on a permanent basis. ]
3. This article was written over the last week, but the breaking news is that Bron Suchecki has received what appears to be the final word on the matter regarding the app gold bars count, which appears to be a standard response sent to other enquirers on the matter.
Dear Ms Suchecki,
You are correct. The number of bars mentioned in the app cannot be used to infer a change in the amount of custodial gold held by the Bank of England as the figure is deliberately non-specific. The Bank will not be offering any further comment on this matter.
Head of Division, Public and Internal Communications Division
Bank of England | Threadneedle Street | London EC2R 8AH
** The app itself apparently allows content update and addition too, so (admittedly) it appears that the text can be whatever the bank wants to set it to, but my main point is that the COPY for the app would have been written in 2012, when the business requirement documents were submitted and signed off. At some point in the development cycle, requirements get locked although subsequent change requests can occur. It was developed as a cross-platform application by agency Clay Interactive in London.
[ Update - 2nd August 2013 ]
On Gene Arensburg's 'gotgoldreport', I saw a good counter argument by 'Simon S' who points out the text (in virtual tour gold vault section 3) states a specific date. He is correct, and I am embarrassed to have overlooked this detail when doing my review (although I did cover the editing possibility). Here is the text:
... but if I may counter the counter; the new information does not have the effect of clarification. Here's my thoughts:
- The text DOES prove that at least some of the app copy was updated in June 2013.
We could assume that the other text was also reviewed for consistency in the same edit but we can't prove that. There is still a 12-month range during which the primary copy could have been produced. Because only that paragraph is date specific there is no way of knowing (or proving) which parts of the text were (or weren't) updated during the June edit; this underscores my point regarding limit-of-reading error.
- The fact they have a specific date reference on the gold bar value but not on the 400,000 bars figure indicates it was conscious decision to not make the 400,000 date specific - i.e. if they did mean it to be accurate then they would have also appended “in June 2013” to the 400,000 figure. This is consistent with the BoE statement on the matter (regarding the usage of the figure).
Congratulations to Simon S for having eyes sharper than mine. I did use the opportunity to check any other month/year specific references in the rest of the application (couldn't find any). Also, recommend having a look at the 'cash vaults' section, you'll see other vault space which looks similar to the display vaults (but more modern).